Simulation data processing (CAT-3B)

Setup

Packages and functions

library(dplyr, warn.conflicts = FALSE)
library(ggplot2)
library(ggiraph)
library(purrr)

data_identifier = "CAT-3B"

text_as_fun <- function(function_string){
  eval(parse(text = function_string))
}
  
file_names <- list.files("../data/simulation_exports/", pattern = data_identifier)
file_paths <- list.files("../data/simulation_exports/", pattern = data_identifier, full.names = TRUE)

sim_list <- purrr::map(
  .x = file_paths, 
  .f = readRDS
)

Pre-processing

1. Define variables for identifying unique cells

2. Remove all intercepts (not of interest)

3. Remove “gb1b” term - not manipulated on heteroscedasticity

4. Merge into a df

Summaries

Model fit

Power/false-positives

Coverage of CIs

CI width

Bias in beta

Mean difference

SD of estimates

Shifted coverage

Model errors